Transcript of RESEARC SECT
471NOV/DEC 2014—VOL. 69, NO. 6JOURNAL OF SOIL AND WATER
CONSERVATION
RESEARCH SECTION
Andrea D. Basche is a graduate research assis- tant in the
Department of Agronomy, Iowa State University in Ames, Iowa.
Fernando E. Miguez is an assistant professor in the Department of
Agronomy at Iowa State University in Ames, Iowa. Thomas C. Kaspar
is a research scientist at the USDA Agricultural Research Service
National Laboratory of Agriculture and the Environment in Ames,
Iowa. Michael J. Castellano is an assis- tant professor in the
Department of Agronomy at Iowa State University in Ames,
Iowa.
Do cover crops increase or decrease nitrous oxide emissions? A
meta-analysis A.D. Basche, F.E. Miguez, T.C. Kaspar, and M.J.
Castellano
Abstract: There are many environmental benefits to incorporating
cover crops into crop rotations, such as their potential to
decrease soil erosion, reduce nitrate (NO3) leaching, and increase
soil organic matter. Some of these benefits impact other
agroecosystem processes, such as greenhouse gas emissions. In
particular, there is not a consensus in the literature regarding
the effect of cover crops on nitrous oxide (N2O) emissions.
Compared to site-specific studies, meta-analysis can provide a more
general investigation into these effects. Twenty-six peer-re-
viewed articles including 106 observations of cover crop effects on
N2O emissions from the soil surface were analyzed according to
their response ratio, the natural log of the N2O flux with a cover
crop divided by the N2O flux without a cover crop (LRR). Forty
percent of the observations had negative LRRs, indicating a cover
crop treatment which decreased N2O, while 60% had positive LRRs
indicating a cover crop treatment which increased N2O. There was a
significant interaction between N rate and the type of cover crop
where legumes had higher LRRs at lower N rates than nonlegume
species. When cover crop residues were incorporated into the soil,
LRRs were significantly higher than those where residue was not
incorporated. Geographies with higher total precipitation and
variability in precipitation tended to produce higher LRRs.
Finally, data points measured during cover crop decompo- sition had
large positive LRRs and were larger than those measured when the
cover crop was alive. In contrast, those data points measuring for
a full year had LRRs close to zero, indicating that there was a
balance between periods when cover crops increased N2O and periods
when cover crops decreased emissions. Therefore, N2O measurements
over the entire year may be needed to determine the net effect of
cover crops on N2O. The data included in this meta-analysis
indicate some overarching crop management practices that reduce
direct N2O emissions from the soil surface, such as no soil
incorporation of residues and use of non- legume cover crop
species. However, our results demonstrate that cover crops do not
always reduce direct N2O emissions from the soil surface in the
short term and that more work is needed to understand the full
global warming potential of cover crop management.
Key words: cover crops—global warming
potential—meta-analysis—nitrous oxide
Agricultural soils account for 69% of nitrous oxide (N2O) emissions
in the United States (USEPA 2013). This occurs because nitrogen (N)
is an essential nutri- ent for agricultural production; N is added
to soil as N fertilizer and manure, released from soil organic
matter, and has high reac- tivity and mobility in terrestrial
ecosystems (Robertson and Vitousek 2009). Fertilizer N recovery
efficiency for major cereal pro- duction is less than 50% and even
as low as 20% (Cassman et al. 2002), which potentially makes large
quantities of N available for the biological processes that release
N2O. Nitrous oxide, which has 300 times the radiative
forcing per mass unit compared to carbon dioxide (CO2), has been
calculated to be the largest contributor to global warming
potential from agricultural cropping systems (USEPA 2013; IPCC
2007; Robertson et al. 2000). Therefore, small reductions in N2O
emissions from agricultural soils can have an overall large impact
on global warming potential. The challenge is to find agricul-
tural management practices with consistent reductions in N2O
emissions across locations, cropping systems, and years given the
high spatial and temporal variability of emissions (Venterea et al.
2012).
Emissions of N2O from terrestrial ecosys- tems are a function of
available mineral N, soil water content, the availability of
electron donors (such as labile carbon [C]), and soil physical
properties (Davidson et al. 2000; Firestone and Davidson 1989;
Venterea et al. 2012). Cover crops may impact aspects of all these
processes in ways that could potentially increase or decrease N2O
emissions as is out- lined in table 1. For example, a growing cover
crop can decrease soil mineral N by incor- porating it into its
biomass, while a legume cover crop may increase soil mineral N via
N fixation (Kaspar and Singer 2011). While alive, cover crops can
decrease soil water through transpiration. After termination, the
mulching effect of cover crop residues on the soil surface may
increase soil water and the potential for denitrification depend-
ing upon timing of precipitation (Dabney 1998). Additionally,
decomposing cover crop residues can temporarily immobilize soil N
and then later increase soil pools of labile C and inorganic N
(Kaspar and Singer 2011; Steenwerth and Belina 2008), which will
also impact dynamics of N2O emissions.
There are many well-researched benefits to incorporating cover
crops into crop rotations, such as their potential to decrease soil
erosion, reduce nitrate (NO3) leaching, increase soil organic
matter, reduce pest and weed pressure, and provide additional soil
N for cash crops (Kaspar and Singer 2011; Doran and Smith 1991).
However, the net impact of cover crops on N2O is not well
understood (Cavigelli et al. 2012; Cavigelli and Parkin 2012).
Although cover crops may temporarily decrease soil NO3 pools and
leaching losses, C can be the substrate limiting N2O emissions in
some agroecosystems. In these situations, a cover crop’s
contribution to the labile C pool can enhance N2O emissions from
the soil surface (Mitchell et al. 2013).
Meta-analysis is an approach that can be used to improve
understanding of the fac-
doi:10.2489/jswc.69.6.471
w
ater C onservation
Table 1 Drivers of nitrous oxide (N
2 O) loss and potential influential factors investigated in
the
meta-analysis. A full description of database variables appears in
table 3.
Denitrification driver Database factor
Mineral nitrogen (N) C:N residue ratio Type of cover crop
Incorporation of residue N fertilizer rate Tillage Reactive carbon
(C) Soil organic C Biomass input from cover crop Type of cover crop
Incorporation of residue Tillage Soil water Biomass input from
cover crop Precipitation Drainage Soil physical properties Bulk
density Soil texture
tors affecting N2O emissions through the systematic review and
quantitative summary of effect size from individual studies. Many
studies investigating cover crops and N2O are conducted on short
time scales (≤2 years) under specific management and climate con-
ditions which may make it difficult to detect differences.
Meta-analysis allows these studies to be pooled and the factors
affecting N2O emissions investigated. The effect of other
conservation practices on N2O emissions have been similarly
evaluated using meta-ana- lytic methods (Six et al. 2004; Van
Kessel et al. 2013), but none to our knowledge that have used
meta-analysis to examine the existing lit- erature on cover crop
effects on N2O.
The objectives of this study were to use a meta-analysis approach
to (1) examine the relative impact of cover crops on N2O emissions
and (2) determine what manage- ment and environmental factors
contribute to variability in cover crop effects on N2O emissions.
There were several factors that we hypothesized would have a large
contribution to this variability. First, we hypothesized that the
type of cover crop (legume versus nonlegume) would have different
effects on N2O emissions; namely legumes would have a greater
potential to increase N2O emissions versus nonlegumes. Second, we
hypothesized that precipita- tion and cover crop biomass would
impact N2O emissions because denitrification also requires
anaerobic conditions and C. Finally, we hypothesized that the
timing of measure- ments was influential in how cover crops impact
N2O, namely that the period imme-
diately following cover crop termination and the subsequent
decomposition would have the largest N2O emissions because of N and
C release from residues.
Materials and Methods Database Development. For the purposes of
this study, we defined a cover crop as a plant not intended to be
harvested, that is grown during a fallow period between harvest and
planting of two cash crops. This included treatments labeled as
cover crops, green manures, or catch crops. A literature review
utilizing electronic databases Google Scholar and Web of Science
was conducted with the following search string: “nitrous oxide
emis- sions or greenhouse gas emissions and cover crops or green
manures or catch crops.” This combination of key terms resulted in
approx- imately five thousand papers. To reduce the number of
papers included in the meta-anal- ysis, the following criteria were
applied: 1. Studies in which the cover crop is not
harvested and is grown between the harvest and planting of cash
crops.
2. Studies reporting N2O measurements. 3. Studies with a control
treatment varying
only in the inclusion of a cover crop and keeping all other
management practices such as tillage and N additions equal.
4. Studies that provided enough information (standard errors,
standard deviations, co- efficients of variation, etc.) about
experi- mental error either in the published paper or in
information that was provided by the authors when contacted to
allow for an estimate of within study variance.
5. Studies published before December of 2012. On the basis of these
criteria, 26 peer-re-
viewed studies representing 19 field experiments (83 observations),
2 growth chamber studies (9 observations), and 5 modeling
experiments with validation data (14 observations) were selected
for inclusion in a database (table 2) (n = 106 observations).
We omitted studies measuring emissions from cover crop treatments
where the cover crop was not grown in the soil on which the
measurements were taken (Bhattacharyya et al. 2012; Petersen et al.
2013). We also omitted papers analyzing emissions of varied
cropping rotations if they did not have a true control treatment
aligning with the cover crop treat- ment, as these would not allow
for a proper comparison (Liebig et al. 2010; Gomes et al. 2009). If
an experimental design matched our criteria, but the publication
did not include enough detail to perform required calcula- tions,
authors were contacted when possible to obtain this
information.
Data Analysis. Environmental and management factors were included
in the database to examine factors that might be correlated with
variabilility among obser- vations. The full list of these factors
is summarized in table 3 and describes categor- ical vs. numeric
variables and the number of observations included in each analysis.
For some of the factors, information that was not directly
available in the studies was derived from other sources and is
described below.
Precipitation. Unless the rainfall data was explicitly reported by
the experi- ments, National Oceanic and Atmospheric
Administration's (NOAA) Global Historical Climatology Network-Daily
database was utilized (Menne et al. 2012) from the closest
available stations over the specific range of dates when N2O was
sampled.
Soil Properties. Reported values for soil texture (percent of sand,
silt, and clay), pH, organic C, and drainage class categoriza- tion
were directly included in the database. If these values were not
reported, the Web Soil Survey (Soil Survey Staff 2012) or litera-
ture for experiments conducted on the same fields was utilized.
Drainage for non-US sites was determined either via contacting
indi- vidual authors or by soil classification. Soil classification
was determined by the refer- enced literature, and all sites were
converted to one of the World Reference Base Group and US Soil
Classification Group equivalents using Krasilnikov et al.
(2009).
C opyright ©
w
ater C onservation
Table 2 Summary of studies included in the meta-analysis.
Cash crop(s) Cover crop(s) Location Reference
Oats Nonlegume and legume Scotland, UK Baggs et al. 2000 Corn
Nonlegume and legume Maryland, USA Rosecrance et al. 2000*
Rice–wheat Legume Ludhiana, India Aulakh et al. 2001 Rice Legume
Jiangxi, China Xiong et al. 2002 Wheat–corn Nonlegume and legume
England, UK Baggs et al. 2003 Corn Nonlegume England, UK
Sarkodie-Addo et al. 2003 Corn Legume Nyabeda, Kenya Millar et al.
2004 Barley Nonlegume Foulum, Denmark Olesen et al. 2004† Soybean
Nonlegume Iowa, USA Parkin et al. 2006* Corn–soybean Nonlegume
Iowa, USA Parkin and Kaspar 2006 Corn–soybean Nonlegume Illinois,
USA Tonitto et al. 2007† Corn–soybean Nonlegume and legume Iowa,
USA Farahbakhshazad et al. 2008† Corn–soybean Nonlegume Michigan,
USA Fronning et al. 2008 Grapes Nonlegume California, USA
Steenwerth and Belina 2008 Rice Nonlegume and legume Kanto Plains,
Japan Zhaorigetu et al. 2008 Corn–pasture–alfalfa Nonlegume
Pennsylvania, USA Chianese et al. 2009† Corn–soybean Nonlegume
Iowa, USA Jarecki et al. 2009 Corn silage Legume Turin, Italy
Alluvione et al. 2010 Corn–tomato, Nonlegume and legume California,
USA De Gryze et al. 2010† Tomato–cotton, Legume Tomato–safflower–
corn–wheat Tomato Legume California, USA Kallenbach et. al 2010
Corn Nonlegume Michigan, USA McSwiney et al. 2010 Tomato Nonlegume
California, USA Barrios-Masias et al. 2011 Corn Nonlegume New York,
USA Dietzel et al. 2011 Barley Nonlegume Foulum, Denmark Petersen
et al. 2011 Corn–soybean Nonlegume Indiana, USA Smith et al. 2011
Tomato Nonlegume California, USA Smukler et al. 2012 *Growth
chamber experiment. †Model simulation experiment.
Period of Nitrous Oxide Measurement. The included experiments
varied in the length of time and time of year over which N2O
emissions were measured. Thus, we divided the observations based on
the time periods into the following categories: full year, cover
crop growth, cover crop decom- position, and cash crop
growth.
These divisions allowed for an analysis of how cover crops
influence N2O fluxes at dif- ferent times of the year. For full
year, the included observations measured throughout the entire span
of at least one entire year. For cover crop growth, the period
coincided with the time that the cover crop was alive and growing.
In many studies, this aligned with the winter season. For cover
crop decompo- sition, the period coincided with the time of cover
crop termination and potential incorporation into the soil.
Depending upon the design of the experiments, this period
lasted between two weeks at minimum and two months at maximum. This
period often aligned with the spring season as well as fer-
tilization events. For cash crop growth, the period coincided with
the growth of the main cash crop. This period often aligned with
the summer and fall seasons.
The dependent variable was the ratio between the N2O flux with a
cover crop treatment to N2O flux without a cover crop:
RR = N2O emissions cover crop treatment
N2O emissions no cover crop treatment
. (1)
Response ratios (RR) were calculated for all combinations of cover
crop and no cover crop (control) treatments within stud- ies where
these treatment pairs varied solely in the inculsion of a cover
crop. Thus, the number of observations obtained from each
study for the meta-analysis varied according to the study’s
experimental design. Within studies, different cover crop
treatments (fac- torial experiments investigating for example
tillage and cover crops), measurement peri- ods (N2O emissions
reported by season or by individual years), or different species of
cover crops were all counted as individual observations, and
response ratios were deter- mined for each of them.
Then equation 1 was natural log trans- formed (Hedges et al. 1999)
to normalize the data as follows:
LRR = ln RR. (2)
The log ratio ensure that changes in the numer- ator and
denominator are affected equally.
Within study error (Vi) was calculated fol- lowing the method of
Hedges et al. (1999), using reported estimates of variances
and
C opyright ©
w
ater C onservation
2 O).
Description of categorical factors and range for Number of Factor
numerical factors observations
Tillage No-till, conventional tillage 74 Carbon (C):Nitrogen (N)
residue ratio 9 to 48 57 Soil bulk density 1.2 to 2.65 67 pH 5.5 to
8.1 89 Type of cover crop Legume, nonlegume, and biculture 106 N
rate (kg ha–1) 0 to 303 103 Soil incorporation of residues Yes, no
84 Kill date Days between cover crop termination and cash crop
planting (1 to 25) 71 Percentage of Sand 8 to 80 106 Percentage of
Silt 11 to 73 106 Percentage of Clay 5 to 45 106 Percentage of
organic C 0 to 30 cm 0.38 to 2.1 97 Cover crop biomass (kg ha–1)
280 to 14,400 65 Total precipitation (mm) 11 to 906 77 Standard
deviation precipitation (mm) 0.5 to 40 77 Drainage Well-drained,
poorly-drained 69 Period of measurement Full year, cover crop
growth, cover crop decomposition, cash crop growth 80 Experiment
type Field, model, growth chamber 106
converting to standard deviations based on experimental
replications:
Vi = + SDcc2
ncc × ycc2
, (3)
where SDcc is the standard deviation of the cover crop treatment,
ncc is the replications of the cover crop treatment, ycc is the
mean N2O emissions of the cover crop treatment, and ncc represents
the N2O emissions of the control or no cover crop treament.
Equation 3 assumes that reported means are normally
distributed.
The first step of the analysis was to deter- mine if there was
homogeneity among the LRR values from all the studies in the data-
set (Hedges and Olkin 1985; Miguez and Bollero 2005). This tested
the assumption that all of the LRR values came from the same
population. If the test was significant, the effect of cover crops
varied among obser- vations, and other factors were affecting the
response. If the test was not significant, then we could conclude
that the cover crops had a similar effect across
observations.
An inverse variance weighting factor (Wi) was used in this step to
weigh each of the 106 LRR values, where studies with larger vari-
ances were weighted less heavily in the analysis. This is one way
by which we can account for the assumed unequal variances among
stud-
ies (Hedges et al. 1999). The inverse variance weighing factor is
calculated as
Wi = 1/Vi . (4)
In the next step of the analysis, mixed model regression analyses
were conducted to individually examine the relative effects of each
of the 18 environmental and man- agement factors on LRR (the
natural log of response ratio) while accounting for the variation
between studies (St-Pierre 2001) with the weighting factor
(equation 4). The database’s environmental and management factors
were treated as fixed effects while study and intercept were
treated as random effects. The statistical model used was
Lij = βo + si + β1Aij + biAij + eij. (5)
where Lij is natural log of the response ratio of ith study,
receiving jth level of fixed factor A (factors in the analysis
[table 3]). βo is the overall intercept across all studies, si is
the ran- dom effect due to the ith level of study (i = 1,…,26), β1
is the fixed regression coefficient of Li on A across all studies,
bi is random effect of study i on the regression coefficient β1,
and eij is the residual error. This general model was first used to
test each of the 18 factors individ- ually. In these analyses, the
N rate factor was found to have the largest effect on Lij. Next, a
second series of regression analyses were per-
formed using models with the N rate factor plus one of the other 17
factors and its inter- action with N rate. The statistical analysis
was performed using the MIXED procedures of SAS (SAS Institute
2010).
For studies that simultaneously mea- sured changes to NO3 leaching,
response ratios were generated to estimate the effect of the cover
crop on these N fluxes. These response ratios represent the natural
log of NO3 leaching in the study’s cover crop treat- ment divided
by the measured value from the no cover crop treatment. When
analyzed alongside the N2O LRR values created in the same manner,
these values provide a more complete understanding of a cover
crop’s role in these parts of the N cycling.
Finally, a sensitivity analysis was performed in order to test the
robustness of the data- base and overall conclusions. We repeated
the homogeneity test and mixed model regres- sion analyses
excluding all individual field and growth chamber studies one at a
time as well as for a subset of the data excluding all of the
modeling studies (Tudoreanu and Phillips 2004; Philibert et al.
2012). This provided an indication of whether the dominant factors
were still significant as the database changed.
Results and Discussion Overall. A test of homogeneity for the data
set was significant (entire data set p = <0.0001, excluding
modeling studies in
C opyright ©
w
ater C onservation
475NOV/DEC 2014—VOL. 69, NO. 6JOURNAL OF SOIL AND WATER
CONSERVATION
sensitivity analysis p = <0.0001), indicating that the LRRs
varied significantly among observations. This means that the effect
of cover crops varied among the data points in our analysis and
that other factors were affecting the response. Forty percent of
the studies assessed in this analysis showed that cover crops
decreased N2O emissions (neg- ative LRR) and 60% of the studies
showed that cover crops increased N2O emissions (positive LRR
[figure 1]). To analyze these general trends, other factors that
potentially affect N2O emissions are discussed separately. Table 4
presents the results of the regression analysis of factors
affecting the LRR includ- ing regression coefficients for the
continuous variables. Positive coefficients indicate that LRR
increases with increases the indepen- dent variable, while negative
coefficients indicate that the LRR decreases with increases in the
independent variable.
Nitrogen Rate. It is well documented that higher N rates increase
N2O emissions (Eichner 1990; Bouwman et al. 2002; Stehfast and
Bouwman 2005). Our statistical analyses evaluating management and
environmental factors revealed that N rate explained more of the
LRR variability than other factors (table 4). In the sensitivity
analysis, N rate was signifi- cant (at the p <0.0001 level) when
excluding the modeling experiments and in 100% of the regression
analyses when excluding each of the 19 field and 2 growth chamber
studies. As a result, interactions with N rate and other factors
were investigated.
There was a significant interaction between the type of cover crop
and N rate (figure 2). When no additional N is applied (zero N
application rate), legumes exhibited higher LRRs than nonlegume
species. This is con- sistent with the results of Gomes et al.
(2009) who found that legume cover crop residues, which have C:N
ratios less than 25, stimulated N mineralization rates in maize
(Zea mays L.) systems with no additional N applications. Because a
significant quantity of mineralized N is subsequently nitrified,
this may enhance NO3 substrate for N2O production. In a lab-
oratory incubation experiment, Huang et al. (2004) observed that
low C:N crop residue ratios increased N2O emissions. Consistent
with the negative relationship between crop residue C:N ratios and
N2O emissions in the absence of additional N inputs, nonlegume
cover crops showed a slight increase in LRRs as N fertilizer rate
increased, reflecting the
Figure 1 Natural log of response ratios (LRR) for 106 observations
in the dataset, where the response ratio represents the nitrous
oxide (N
2 O) flux with a cover crop divided by the N
2 O flux without a
cover crop.
O bs
er va
tio n
ra nk
1 to
1 06
LRR
Table 4 F and p values for all environmental and management factors
in the mixed model regression analysis. Regression coefficients are
presented for the continuous variables analyzed. DF = degrees of
freedom.
Error Regression Source DF DF coefficient F value Pr > F
Tillage 1 54 2.7 0.106 Carbon (C): nitrogen 1 43 –0.04 2.17 0.1483
(N) residue ratio Soil bulk density 1 51 0.97 2.7 0.1063 pH 1 65
0.68 15.57 0.0002 Type of cover crop 2 78 2.51 0.0878 N rate 1 77
0.00 364.58 <0.0001 Soil incorporation 1 64 5.84 0.0186 Kill
date 1 53 –0.03 1.14 0.2901 Percentage of sand 1 79 0.36 0.36
0.5494 Percentage of silt 1 79 –0.24 0.12 0.7297 Percentage of clay
1 79 –1.23 0.65 0.4217 Percentage of organic carbon 1 74 –0.56 4.05
0.0478 Cover crop biomass 1 49 0.00 0.74 0.3947 Total precipitation
1 58 –0.00 8.49 0.0051 Standard deviation 1 58 0.11 10.66 0.0018
Precipitation drainage 1 54 0.03 0.8693 Period of measurement 3 57
54.94 <0.0001 Experiment type 2 80 0.73 0.4862
C opyright ©
w
ater C onservation
importance of both C and N for the denitri- fication process.
There was also a significant interaction between N rate and tillage
system (figure 3). Mechanical soil disturbances have been observed
to stimulate C mineralization and net N mineralization (House et
al. 1984; Beare et al. 1994; Omonode et al. 2011) due to the
disruption of soil aggregates which expose organic C to microbial
decompo- sition. In no-till systems, LRRs slightly increased with
increasing N rate. This may have occurred because increasing cover
crop biomass on the soil surface with increasing N fertilizer rate
could have mulched the soil surface keeping it slightly wetter. In
conven- tionally tilled systems, lower N rates tended to result in
positive LRRs. This suggests that at higher N rates in a
conventionally tilled system, the cover crop may contribute to a
reduction in N2O emissions relative to the control treatment
without cover crops.
Further, even negative LRRs (cover crop treatments reduced N2O) may
not reflect a large reduction in the overall magnitude of N2O
emissions, particularly with high N fertilization rates. Table 5
includes a subset of studies reporting N2O in kg ha–1 (LRRs were
generated using the reported units which varied by study and the
length of measurement) to demonstrate the magnitude of changes with
and without cover crops. Cover crops reduced N2O emissions at high
N rates (~1 to 2 kg [2.2 to 4.4 lb] N2O dif- ference in study 1 and
2) or by a neglibile amount at 0 N rates (study 3). In other stud-
ies, cover crops increased N2O emissions by 2 to 4 kg ha–1 (1.8 to
3.6 lb ac–1)at higher N rates (study 4 and 5). Finally, study 6
indi- cated a large increase (~40 kg [88.1 lb] N2O) in N2O
emissions at a 0 N rate, given the large N contribution from a
legume cover crop and the anaerobic soil conditions in the cropping
system. Further, this large release of N2O occurred while the cover
crop was decomposing, a period observed to have high N2O emissions
(figure 5).
Type of Cover Crop. Cover crops were categorized into the following
types: legume (such as clover, vetch [Vicia villosa], field bean,
and pea varieties), nonlegume (such as cereal rye [Secale cereal],
annual ryegrass [Lolium mul- tiflorum], oats [Avena sativa], wheat
[Triticum aestivum], and radish mustards), and bicul- ture species
(such as vetch and rye mixes). In general, legumes typically
resulted in pos- itive LRRs, while the LRRs for nonlegume
Figure 2 Response ratios (LRR) of legume versus grass cover crop
species as a function of fertilizer nitro- gen (N) rate. At the 0 N
rate, legume cover crops have a higher response ratio than grass
cover crop species. Across a range of N application rates, the
response ratio for nonlegume cover crop species only increases
slightly; for legumes the trend declines.
LR R
N Rate (kg N ha–1)
1
0.5
0
Legend Nonlegume Legume
and biculture species were close to zero (fig- ure 4). Statistical
analysis revealed that there was a significant difference at the p
<0.10 level in response ratios between the legume, cover crop
type nonlegume and biculture groups. In the sensitivity analysis
excluding the five modeling studies, type was found to be
significant (p = 0.002), and we thus cannot reject our hypothesis
that cover crop type influences cover crop impact on N2O emissions.
Because cover crops take up N that might otherwise be lost to
leaching or because legume cover crops can fix N, cover crops may
increase soil N availability during decomposition and, thus, may
increase the available NO3 substrate for denitrification and N2O
emissions within agricultural fields.
Period of Measurement. Based on the period of measurement, cover
crops influ- enced N2O dynamics differently throughout the year (p
<0.0001). The sensitivity analysis further revealed that period
of measurement was significant (at the p <0.05 level) in 95% of
the statistical models when excluding individual studies. Data
points based on mea- surements made across an entire year had an
average response ratio close to zero com-
pared to the other periods of measurement (figure 5). This may
suggest that there is a net neutral effect of a cover crop on N2O
emis- sions when measured over longer timescales. Figure 5
illustrates that even if particular periods of the year see larger
N2O impacts of a cover crop, a full-year time scale may actu- ally
find a net neutral effect. More long-term field experiments
measuring N2O over the entire year are needed to better understand
these dynamics.
Our analysis indicated that the highest LRRs were data points
measuring during the cover crop decomposition period, con- sistent
with our hypothesis. Rosecrance et al. (2000) observed the largest
N2O fluxes over the course of a growth chamber experiment in the
five days after cover crop termina- tion with rye, vetch, and a
mixture of both (C:N of 21, 10, and 14 respectively). They
concluded that additional C substrate plus available mineral N
contributed to high N2O emissions during this period. Aukulah et
al. (2001) also found that N2O production was highest in the
initial four week period fol- lowing legume cover crop soil
incorporation in a flooded rice system. They attributed this
C opyright ©
w
ater C onservation
Figure 3 Response ratios (natural log of nitrous oxide [N
2 O] flux with a cover crop divided by the N
2 O flux
without a cover crop [LRR]) of conventionally tilled and no-tilled
systems as a function of nitro- gen (N) application rate. Cover
crops reduced response ratios at higher N rates in conventionally
managed systems. No-till systems increased response ratios slightly
(compared to conventional tillage) as N rates increased.
LR R
N Rate (kg N ha–1)
0.5
0
–0.5
Table 5 Magnitude of nitrous oxide (N
2 O) changes with and without cover crops for subset of data base
studies.
No cover Cover crop Nitrogen (N) crop N2O N2O application Study
emissions emissions Cropping system and Measurement rate number (kg
N ha–1) (kg N ha–1) cover crop species period (kg ha–1)
Reference
1 7.5 5.3 Corn in corn—soybean, 70% rye/ 30% oat Full year 175
Jarecki et al. 2009 2 3.7 2.3 Soybean, winter rye Winter (cover 195
Parkin et al. 2006 crop growth) 3 1.5 1.4 Soybean in corn—soybean,
annual ryegrass Full year 0 N Smith et al. 2011 4 11.3 15.4 Corn in
corn—soybean, winter rye Full year 215 Parkin and Kaspar 2006 5 3.8
5.1 Corn in corn—soybean, annual ryegrass Full year 193 Smith et
al. 2011 6 9.3 50.2 Rice—wheat, sesbania Spring (cover 0 N (176
Aulakh et al. 2001 crop decomposition) from legume CC)
to the interaction between NO3 and organic C availability, given
that soil water con- tent and temperature remained consistently
favorable for denitrification. Sarkodie-Addo et al. (2003) measured
NO3, ammonium (NH4), and N2O for 55 days after incorpo- ration of a
wheat and winter rye cover crop with and without fertilizer.
Fertilized plots had positive LRRs, and nonfertilized plots
had negative LRRs. They reported that the decrease in N2O emissions
with cover crops in the nonfertilized plots could be a result of
temporary N immobilization from the cover crop’s C contribution.
The results of the studies measuring N2O during the cover crop
decomposition period suggest that N2O emissions are affected by the
interaction of C input and N availability. Cover crop resi-
dues with low C:N ratios generally increased N2O emissions
(positive LRR [figure 6]) during the decomposition period. This is
consistent with observations of Millar et al. (2004) that N2O from
systems with legume cover crops were positively correlated with
residue N content. Further, the positive LRR observed during the
growth of the cash crop may indicate that there is still some cover
crop decomposition happening during this period.
Studies during the growth of the cover crop period had the lowest
mean LRR of all the periods of measurement (figure 5). This could
be a result of cover crop N uptake as well as the fact that this
period often occurred during the winter when temperatures are
lower. Temperature is important because microbial process rates
including N miner- alization, nitrification, and denitrification
exponentially decline with decreasing tem- perature (Stanford et
al. 1975). In a growth chamber study, in which temperature was
controlled, Parkin et al. (2006) found that winter rye cultivated
with manure treatments reduced available soil NO3 as well as N2O
emissions compared with levels measured in the no cover crop
treated pots. This suggests that crop N uptake creates a larger
sink for the soil mineral-N pool than N2O emis- sions or NO3
leaching. Dietzel et al. (2011) measured N2O emissions in a
maize-winter rye cover crop system over two winter and spring
seasons. The two years varied signifi- cantly in winter conditions
which altered the soil water status by changing the frequency of
freezing and thawing cycles. The warmer winter resulted in more
negative LRRs than the colder winter when more freeze
C opyright ©
w
ater C onservation
Ty pe
LRR
Yes
No
Biculture
Legume
Nonlegume
n = 30
n = 4
n = 33
n = 69
Figure 5 Mean response ratios (LRR; 95% confidence intervals also
shown) for environmental factors included in the meta-analysis: the
period of measurement, the total precipitation over the measurement
period, and the standard deviation of precipitation over that
period.
Pe rio
LRR
>10mm
SD p
re ci
pi ta
tio n
n = 17
n = 24
n = 36
n = 24
n = 22
n = 31
n = 7
n = 40
n = 11
n = 22
Pr ec
ip ita
tio n
thaw cycles were present. In this study, the cover crop response
ratio’s dependence on weather variability may further illustrate
the value of measuring over multiple seasons or years (larger time
scales) to better understand annual cover crop N2O dynamics.
Soil Incorporation. In our analysis, LRRs for studies that
incorporated cover crop res- idues into the soil were significantly
higher than those for studies that left the residues on the soil
surface (p = 0.02 [figure 4]). Of the studies where incorporation
was reported, 19 of the 20 highest positive response ratios in the
database were cover crop treatments where residues had been
incorporated into the soil. The sensitivity analysis also found
that soil incorporation was significant (p < 0.05) in 81% of the
models when exclud- ing individual studies. Incorporation of cover
crop residues contributes to an increase in N2O emissions through
several potential effects. Incorporation of cover crop residues
increases N mineralization rates of both soil organic matter and
cover crop residues, and it contributes to greater NO3 availability
and denitrification (Firestone and Davidson 1989). Incorporation of
cover crops residues also likely increases soil temperature and,
thus, the potential for denitrification compared with soil covered
with residues (Omonode et al. 2011). Lastly, anaerobic conditions
for denitrification of cover crop N is more likely to occur if the
residues are incorporated with tillage rather than left on the
surface (Kaspar and Singer 2011). Thus, our analysis indi- cated
that incorporating aboveground cover crop residues led to relative
increases in N2O emissions through a variety of mechanisms.
Precipitation. The episodic nature of N2O emissions results in part
from the requirement for denitrification for anaerobic soil condi-
tions, which usually occur following large or intense precipitation
events (Davidson et al. 2000). Cover crops may alter the soil water
status and the potential for anaerobic condi- tions in several
ways, including decreased soil evaporation, increased rainfall
infiltration, and transpiration of stored soil water during cover
crop growth (Unger and Vigil 1990). To evaluate the soil water
status and potential for anaerobic condition of a study, we
utilized total precipitation over the measurement period as well as
the standard deviation of the rainfall as indicators for conditions
favoring development of anaerobic soil conditions and
denitrification. Similarly, the DeNitrification- DeComposition
(DNDC) model (Li et al.
C opyright ©
w
ater C onservation
479NOV/DEC 2014—VOL. 69, NO. 6JOURNAL OF SOIL AND WATER
CONSERVATION
Figure 6 Response ratios (LRR) for observations measured during the
cover crop decomposition period as a function of the residue carbon
(C):nitrogen (N) ratio. Legume species and those species with lower
C:N ratios frequently led to an increase in nitrous oxide (N
2 O) emissions, as indicated
by the positive response ratios.
LR R
3
2
1
0
–1
Legend Nonlegume Legume
1992) uses daily precipitation along with other variables as a
predictor of the N2O emissions. Other models like Agricultural
Production Systems Simulator (APSIM) (Thorburn et al. 2010) use
water filled pore space as a predic- tor of N2O emissions.
In the statistical model testing the effect of precipitation values
on LRRs, total precipi- tation (p = 0.005) and the standard
deviation of precipitation (p = 0.002) were significant (figure 5).
As we hypothesized, precipitation is an important factor impacting
the LRRs. Total precipitation, however, was significant at the p
< 0.1 level in 86% of the statistical models excluding
individual studies, while the standard deviation of precipitation
was significant at the p < 0.05 level in 95% of the statistical
models. Studies with legume cover crops had a more pronounced trend
toward increased response ratios as the total precipi- tation and
standard deviation of precipitation increased. All of the
observations (20 points representing 7 different studies, where 77
total points were included in this part of the analysis) with a
standard deviation of precip- itation above 8.8 mm (0.34 in) had
positive LRRs. This may indicate that regardless of
other factors (such as cover crop type) above a threshold of
rainfall variability, a cover cropped agroecosystem is more
susceptible to N2O emissions than one without a cover crop. Novoa
and Tejeda (2006) noted that N2O emissions from applied plant
residues were predicted in part by rainfall. This could be a result
of a cover crop residue maintain- ing higher soil moisture and
providing labile C, along with the timing of high intensity
rainfall events.
Soil Organic Carbon. Soil organic C (SOC) has a strong impact on N
transfor- mations including the denitrification process (Davidson
et al. 2000). In addition, many models (APSIM, Daily Century Model
[DAYCENT], DNDC, Environmental Policy Integrated Climate Model
[EPIC], and ecosys) capable of simulating N2O emis- sions include
SOC as a predictor (Li et al. 1992; Adler et al. 2007; De Gryze et
al. 2010; Thorburn et al. 2010). Cover crops are a source of C and
therefore the amount and quality of additional biomass has the
poten- tial to alter N2O emissions. Two factors were categorized
and analyzed to evaluate the effect of SOC on LRRs: percentage
organic
C in the topsoil and total cover crop biomass. The percentage
organic C of the topsoil was found to be significant in the
statistical model testing its effect on the LRR (p = 0.05). With
larger SOC values in the topsoil, the LRR showed a small decline.
Bouwman et al. (2002) found significantly larger N2O emis- sions in
soils with 3% to 6% organic C versus those with 1% to 3%. However,
the experi- ments included in this analysis had a much smaller
range of SOC values (0.38% to 2.10% [table 3]), which may be one
reason we did not observe as clear of a relationship between SOC
and N2O emissions. It is possible that at lower background levels
of SOC, higher LRRs could be a result of a larger cover crop effect
due to C limitation. Additionally, our analysis indicated that the
total amount of cover crop biomass did not have a significant
effect on LRRs, although there was a trend toward higher LRRs as
biomass increased (data not shown). Contrary to our hypoth- esis
that cover crop biomass would be an important factor controlling
N2O emissions we found inconclusive evidence of this. The
sensitivity analysis found cover crop biomass significant at the p
< 0.10 level in 62% of the regression analyses excluding
individual studies. Robinson and Conroy (1999) found that when
elevated CO2 levels increased plant productivity, subsequent
additional C substrate for microbes contributed to con- sumption of
more soil O2 than could be replaced by diffusion. This led to
anerobic soil conditions and increased denitrification. This
mechanism seems consistent with our analysis, given the
relationships in the dataset with LRRs, SOC, cover crop biomass,
and precipitation. It also underscores multiple interconnections
between C and N cycling in agroecosystems.
Cover Crops and Global Warming Potential. Nitrate lost through
leaching from agricultural fields is subject to denitrification and
N2O emissions off-site, which would not be reflected in the on-site
measurements of N2O emissions from the soil surface. Therefore,
given the ability of cover crops to reduce NO3 leaching, cover
crops may con- tribute to an overall decrease in net global warming
potential. Mosier et al. (1998) estimated indirect N2O emissions
result- ing from leaching and runoff to be 2.5% of total leached N.
They further calculated that indirect denitrification (e.g., from
leach- ing and runoff) emissions constitute 25% of global N2O
emissions from agricultural soils.
C opyright ©
w
ater C onservation
Figure 7 The mean nitrate (NO
3 ) leaching response ratios (LRR; natural log of the NO
3 leaching with a
cover crop divided by the NO 3 leaching without a cover crop) and
95% confidence intervals
compared to the mean nitrous oxide (N 2 O) response ratios from
three studies measuring both.
Ten of the 11 points were measured during the cover crop growth
period. Although this rep- resents only a small subset of the data
base, it could further suggest that cover crop nitrogen uptake
during growth decreases leaching losses and subsequent indirect
N
2 O emissions.
LRR
N20
N03 leaching
n = 11
n = 11
For studies measuring leaching losses in this meta-analysis, mean
change in NO3 loss with a cover crop was significantly lower than
the slight increase to neutral effect on direct N2O emissions
(figure 7). This is consistent with the results of Tonitto et al.
(2006) who found that on average nonlegume cover cropped systems
reduced NO3 leaching by 70% and legume cover cropped systems
reduced NO3 leaching by 40%. Even though indirect esti- mates of
N2O emissions are variable, this is an important impact to consider
that would not be included in the LRR for direct emissions used in
our analysis.
One modeling experiment (De Gryze et al. 2010) and two field
experiments (Fronning et al. 2008; Smith et al. 2011) reported net
global warming potentials (GWP) that were neutral or negative
(indicating mitigative potential) when cover crops were present. In
our database, only these three studies included full net GWPs,
measuring change in SOC (or soil CO2 respiration), N2O, and CH4. De
Gryze et al. (2010) found that the net decrease in GWP was
primarily a result of increased SOC storage in cover cropped
systems. More multiyear field trials and mod- eling efforts are
needed to better understand
the long term effect of cover crops on the net GWP of
agroecosystems.
Summary and Conclusions This meta-analysis found that cover crops
increased N2O emissions from the soil sur- face in 60% of published
observations while cover crops decreased N2O emissions from the
soil surface in 40% of observations. There are both environmental
and man- agement factors that modified the impact of cover crops on
N2O emissions, including fertilizer N rate, soil incorporation, and
the period of measurement and rainfall. Legume cover crops had
higher relative N2O emis- sions at low N rates and lower emissions
at high N rates, whereas N2O emissions of nonlegume cover crops
increased as N rate increased. In general, it seems that cover
crops have a greater potential to reduce N2O emissions when
nonlegume species are utilized and cover crop residue is not incor-
porated into the soil. Our analysis also found that cover crops on
average only lead to a small or negligible increase in N2O emis-
sions when measured for time periods of one year or greater. To
understand the full global impact of cover crops on N2O emissions,
more field research with measurements over
extended time periods is needed to examine the temporal component
of N2O emissions. Better accounting for cover crop reductions in
indirect N2O emissions from leached N should also be
considered.
Acknowledgements This research is part of a regional collaborative
proj-
ect supported by the USDA National Institute of Food
and Agriculture (NIFA), Award No. 2011-68002-30190,
Cropping Systems Coordinated Agricultural Project:
Climate Change, Mitigation, and Adaptation in Corn-based
Cropping Systems. Project website: sustainablecorn.org. The
authors appreciate personnel of the Miguez Agroecosystem
Modeling Lab for their inputs on drafts of this project as
well as Stacie Shuler for her work in data organization. We
are also grateful for the additional information and
responses
provided by the following authors: M. Aulakh, M. Cavigelli,
F. Barrios Masias, R. Dietzel, C. Kallenbach, M. Liebig, C.
McSwiney, N. Millar, K. Thelen, S. Smukler, and S. Snapp.
References Adler, P.R., S.J. Del Grosso, and W.J. Parton. 2007.
Life-
Cycle Assessment of Net Greenhouse-Gas Flux for
Bioenergy Cropping Systems. Ecological Applications
17(3):675-691.
Alluvione, F., C. Bertora, L. Zavattaro, and C. Grignani.
2010.
Nitrous ooxide and carbon dioxide emissions following
green manure and compost fertilization in corn. Soil
Science Society of America Journal 74(2):384-395.
Aulakh, M., T. Khera, J. Doran, and K. Bronson. 2001.
Denitrification, N2O and CO2 fluxes in rice-wheat
cropping system as affected by crop residues, fertilizer
N and legume green manure. Biology and Fertility of
Soils 34(6):375-389.
Cook, and G. Cadisch. 2003. Nitrous oxide emissions
following application of residues and fertiliser under zero
and conventional tillage. Plant and Soil 254(2):361-370.
Baggs, E.M., C.A. Watson, and R.M. Rees. 2000. The fate
of nitrogen from incorporated cover crop and green
manure residues. Nutrient Cycling in Agroecosystems
56:153-163.
Cultivar mixtures of processing tomato in an organic
agroecosystem. Organic Agriculture 1(1):17-30.
1994. Aggregate-Protected and Unprotected Organic
Matter Pools in Conventional- and No-Tillage Soils.
Soil Science Society of America Journal 58:787-795.
Bhattacharyya, P., K.S. Roy, S. Neogi, S.P. Chakravorti,
K.S. Behera, K.M. Das, S. Bardhan, and S. Rao. 2012.
Effect of long-term application of organic amendment
on C storage in relation to global warming potential
and biological activities in tropical flooded soil
planted to rice. Nutrient Cycling in Agroecosystems
94(2-3):273-285.
w
ater C onservation
Bouwman, A.F., L.J.M. Boumans, and N.H. Batjes. 2002.
Emissions of N2O and NO from fertilized fields:
Summary of available measurement data. Global
Biogeochemical Cycles 16(4):6-1-6-13.
Agroecosystems, nitrogen-use efficiency, and nitrogen
management. Ambio 31(2):132-40.
P.E. Fixen, R.T. Venterea, A.B. Leytem, J.E. McLain,
and D.B. Watts. 2012. US agricultural nitrous oxide
emissions: context, status, and trends. Frontiers in
Ecology and the Environment 10(10):537-546.
Cavigelli, M.A., and T.B. Parkin. 2012. Cropland Management
Contributions to GHG Flux: Eastern and Central U.S.
In Managing agricultural greenhouse gases: coordinated
agricultural research through GRACEnet to address our
changing climate. ed. M.A. Liebig, A. Franzleubbers, and
R. Follett. Waltham, MA: Academic Press.
Chianese, D.S., C.A. Rotz, and T.L. Richard. 2009. Simulation
of nitrous oxide emissions from dairy farms to assess
greenhouse gas reduction strategies. Transactions of
the American Society of Agricultural and Biological
Engineers 52(4):1325-1335.
and E. Veldkamp. 2000. Testing a conceptual model of
soil emissions of nitrous and nitric oxides. Bioscience
50(8):667-680.
hydrology. Journal of Soil and Water Conservation.
53(3):207-213.
De Gryze, S., A. Wolf, S.R. Kaffka, J. Mitchell, D.E.
Rolston,
S.R.Temple, J. Lee, and J. Six. 2010. Simulating
greenhouse gas budgets of four California cropping
systems under conventional and alternative management.
Ecological Applications 20(7):1805-19.
Dietzel, R., D. Wolfe, and J.E. Thies. 2011. The influence
of winter soil cover on spring nitrous oxide emissions
from an agricultural soil. Soil Biology and Biochemistry
43(9):1989-1991.
Doran, J.W., and M.S. Smith. 1991. Role of cover crops
in nitrogen cycling. In Cover Crops for Clean
Water, ed. W.L. Hargrove. Ankeny, IA: Soil and Water
Conservation Society.
Fertilized Soils: Summary of Available Data. Journal of
Environmental Quality 19:272-280.
Farahbakhshazad, N., D. Dinnes, C. Li, D. Jaynes, and W.
Salas. 2008. Modeling biogeochemical impacts of
alternative management practices for a row-crop field
in Iowa. Agriculture, Ecosystems & Environment
123(1-3):30-48.
basis of NO and N2O production and consumption
in soil. In Exchange of Trace Gases between Terrestrial
Ecosystems and the Atmosphere, eds. M.O. Andreae and
D.S. Schimel. New York, NY: John Wiley & Sons.
Fronning, B.E., K.D. Thelen, and D.H. Min. 2008. Use of
manure, compost, and cover crops to supplant crop
residue carbon in corn stover removed cropping systems.
Agronomy Journal 100(6):1703.
Gomes, J., C. Bayer, F. De Souza Costa, M. De Cássia
Piccolo, J.A. Zanatta, F.C. Vieira, and J. Six. 2009. Soil
nitrous oxide emissions in long-term cover crops-based
rotations under subtropical climate. Soil and Tillage
Research 106(1):36-44.
Hedges, L., and I. Olkin. 1985. Statistical Methods for Meta-
Analysis. New York, NY: Academic Press.
Hedges, L., J. Gurevitch, and P. Curtis. 1999. The meta-
analysis of response ratios in experimental ecology.
Ecology 80(4):1150-1156.
G.W. Langdale. 1984. Nitrogen cycling in conventional
and no-tillage agroecosysterns in the Southern
Piedmont. Journal of Soil and Water Conservation
39(3):194-200.
Huang, Y., J.W. Zou, X.H. Zheng, Y.S. Wang, and X.K.
Xu. 2004. Nitrous oxide emissions as influenced by
amendment of plant residues with different C:N ratios.
Soil Biology & Biochemistry 36(6):973-981.
Basis. http://www.ipcc.ch/publications_and_data/ar4/
wg1/en/ch2s2-10-2.html#table-2-14.
Moorman, J.W. Singer, B.J Kerr, J.L. Hatfield, and
R. Jones. 2009. Cover crop effects on nitrous oxide
emission from a manure-treated Mollisol. Agriculture,
Ecosystems & Environment 134(1-2):29-35.
Kaspar, T.C., and J.W. Singer. 2011. The use of cover crops
to
manage soil. In Soil Management: Building a Stable Base
for Agriculture, eds. J.L. Hatfield and T.J. Sauer. Madison,
WI: American Society of Agronomy and Soil Science
Society of America.
differently depending on type of irrigation. Agriculture,
Ecosystems & Environment 137(3-4):251-260.
Krasilnikov, P., J. Ibanez Marti, R. Arnold, and S. Shoba.
2009.
A Handbook of Soil Terminology, Correlation and
Classification. London: Earthscan.
Li, C., S. Frolking, and T.A. Frolking. 1992. A model of
nitrous oxide evolution from soil driven by rainfall
events: 1. Model structure and sensitivity. Journal of
Geophysical Research 97(D9):9759-9776.
effects on soil carbon and greenhouse gas flux in central
North Dakota. Soil Science Society of America Journal
74(2):358-365.
Miguez, F.E., and G.A. Bollero. 2005. Review of corn yield
response under winter cover cropping systems using
meta-analytic methods. Crop Science 45(6):2318-2329.
McSwiney, C.P., S.S. Snapp, and L.E. Gentry. 2010.
Use of N immobilization to tighten the N cycle in
conventional agroecosystems. Ecological applications
Menne, M.J., I. Durre, R.S. Vose, B.E. Gleason, and T.G.
Houston. 2012. An overview of the Global Historical
Climatology Network-Daily database. Journal of
Atmospheric and Oceanic Technology 29:897-910.
Millar, N., J.K. Ndufa, G. Cadisch, and E.M. Baggs. 2004.
Nitrous oxide emissions following incorporation of
improved-fallow residues in the humid tropics. Global
Biogeochemical Cycles 18(1):1-9.
of mineralizable carbon. Soil Science Society of America
Journal 77:1-9.
Mosier, A., C. Kroeze, C. Nevison, O. Oenema, and S.
Seitzinger. 1998. Closing the global N2O budget:
nitrous oxide emissions through the agricultural
nitrogen cycle inventory methodology. Nutrient
Cycling in Agroecosystems 52:225-248.
by environmental and management factors. Nutrient
Cycling in Agroecosystems 75(1-3):29-46.
and D. Christen. 2004. Effect of climate change on
greenhouse gas emissions from arable crop rotations.
Nutrient cycling in agroecosystems 70:147-160.
Omonode, R.A., D.R. Smith, A. Gál, and T.J. Vyn. 2011. Soil
nitrous oxide emissions in corn following three decades
of tillage and rotation treatments. Soil Science Society of
America Journal 75(1):152-163.
Parkin, T.B., T.C. Kaspar, and J.W. Singer. 2006. Cover crop
effects on the fate of N following soil application of
swine manure. Plant and Soil 289(1-2):141-152.
Parkin, T.B., and T.C. Kaspar. 2006. Nitrous oxide emissions
from corn-soybean systems in the Midwest. Journal of
Environmental Quality 35(4):1496-1506.
Biochemistry 57:706-712.
Munkholm. 2011. Tillage effects on N2O emissions
as influenced by a winter cover crop. Soil Biology and
Biochemistry 43(7):1509-1517.
of the quality of meta-analysis in agronomy. Agriculture,
Ecosystems & Environment 148:72-82.
Greenhouse gases in intensive agriculture: Contributions
of individual gases to the radiative forcing of the
atmosphere. Science 289(5486):1922-1925.
agriculture: Balancing the cost of an essential resource.
Annual Review of Environment and Resources
34(1):97-125.
w
ater C onservation
Robinson, D., and J.P. Conroy. 1999. A possible plant-
mediated feedback between elevated CO2,denitrification
and the enhanced greenhouse effect. Soil Biology &
Biochemistry 31:43-53.
Teasdale. 2000. Denitrification and N mineralization
from hairy vetch and rye cover crop monocultures and
bicultures. Plant and Soil 227:283-290.
Sarkodie-Addo, J., H.C. Lee, and E.M. Baggs. 2003.
Nitrous oxide emissions after application of inorganic
fertilizer and incorporation of green manure residues.
Soil Use and Management 19:331-339.
SAS Institute. 2010. PC Windows Version 9.3. Cary, NC: SAS
Institute Inc.
Six, J., S.M. Ogle, F.J. Breidt, R.T. Conant, A.R. Mosier,
and K. Paustian. 2004. The potential to mitigate global
warming with no-tillage management is only realized
when practised in the long term. Global Change
Biology 10(2):155-160.
D.L. Bucholtz, and D.E. Stott. 2011. Fertilizer and
tillage management impacts on non-carbon-dioxide
greenhouse gas emissions. Soil Science Society of
America Journal 75(3):1070-1082.
Assessment of best management practices for nutrient
cycling: A case study on an organic farm in a
Mediterranean-type climate. Journal of Soil and Water
Conservation 67(1):16-31, doi:10.2489/jswc.67.1.16.
USDA Natural Resources Conservation Service. http://
websoilsurvey.nrcs.usda.gov/.
Stanford, G., S. Dzienia, and R.A. Vander Pol. 1975. Effect
of
temperature on denitrification rate in soils. Soil Science
Society of America Journal 39:867-870.
Stehfest, E., and L. Bouwman. 2006. N2O and NO
emission from agricultural fields and soils under natural
vegetation: summarizing available measurement data and
modeling of global annual emissions. Nutrient Cycling
in Agroecosystems 74(3):207-228.
enhance soil organic matter, carbon dynamics and
microbiological function in a vineyard agroecosystem.
Applied Soil Ecology 40:359-369.
mixed model methodology. Journal of Dairy Science
84(4):741-55.
emissions from diverse Australian sugarcane production
systems. Agriculture, Ecosystems & Environment
intensive cropping systems: A meta-analysis of crop
yield and N dynamics. Agriculture, Ecosystems &
Environment 112(1):58-72.
2007. Application of the DNDC model to tile-
drained Illinois agroecosystems: model comparison of
conventional and diversified rotations. Nutrient Cycling
in Agroecosystems 78(1):65-81.
soya bean plants. Journal of the Science of Food and
Agriculture 84(8):845-852.
Conservation 53(3):200-207.
B. Linquist, and K.J. Van Groenigen. 2013. Climate,
duration, and N placement determine N2O emissions in
reduced tillage systems: A meta-analysis. Global Change
Biology 19(1):33-44.
2013. Inventory of US Greenhouse Gas Emissions
and Sinks: 1990-2011. Washington, DC. http://www.
epa.gov/climatechange/Downloads/ghgemissions/
US-GHG-Inventory-2013-Chapter-6-Agriculture.pdf.
Nelson, K.A. Spokas, B. Pal Singh, C.E. Stewart, A.
Ranaivoson, J. Strock, and H. Collins. 2012. Challenges
and opportunities for mitigating nitrous oxide emissions
from fertilized cropping systems. Frontiers in Ecology
and the Environment 10(10):562-570.
Xiong, Z.Q.,G.X. Xing, H. Tsuruta, G.Y. Shen, S.L. Shi, and
L.J. Du. 2002. Measurement of nitrous oxide emissions
from two rice-based cropping systems in China.
Nutrient Cycling in Agroecosystems 64:125-133.
Zhaorigetu, M. Komatsuzaki, Y. Sato, and H. Ohta. 2008.
Relationships between fungal biomass and nitrous
oxide emission in upland rice soils under no tillage and
cover cropping systems. Microbes and Environments
23(3):201-208.
w
ater C onservation